Bug 117043

Summary: [em] Intel PWLA8492MT Dual-Port Network adapter EEPROM Checksum is Not Valid
Product: Base System Reporter: Pavel Stepanov <wishmaster>
Component: kernAssignee: jfv
Status: Closed Overcome By Events    
Severity: Affects Only Me CC: sbruno
Priority: Normal Keywords: IntelNetworking
Version: 6.2-STABLE   
Hardware: Any   
OS: Any   

Description Pavel Stepanov 2007-10-09 15:40:01 UTC 
When using both ports on a PWLA8492MT adapter, 

Oct  8 23:15:09 vpn1 kernel: em0: The EEPROM Checksum Is Not Valid
Oct  8 23:15:09 vpn1 kernel: em0: Unable to initialize the hardware
Oct  8 23:15:09 vpn1 kernel: em0: The EEPROM Checksum Is Not Valid
Oct  8 23:15:09 vpn1 kernel: em0: Unable to initialize the hardware
Oct  8 23:15:09 vpn1 kernel: em1: The EEPROM Checksum Is Not Valid
Oct  8 23:15:09 vpn1 kernel: em1: Unable to initialize the hardware
Oct  8 23:15:09 vpn1 kernel: em1: The EEPROM Checksum Is Not Valid
Oct  8 23:15:09 vpn1 kernel: em1: Unable to initialize the hardware

message appear, then network card does not work until reboot. But, if
both ports are used it happening just after system boots up.

If only one port used network card works correctly.


// dmesg
Oct  8 23:20:37 vpn1 kernel: em0: <Intel(R) PRO/1000 Network Connection Version - 6.2.9> port 0x4000-0x403f mem 0xee240000-0xee25ffff,0xee200000-0xee23ffff irq 24 at device 1.0 on pci10
Oct  8 23:20:37 vpn1 kernel: em0: Ethernet address: 00:02:a5:4e:8d:b2
Oct  8 23:20:37 vpn1 kernel: em1: <Intel(R) PRO/1000 Network Connection Version - 6.2.9> port 0x4040-0x407f mem 0xee260000-0xee27ffff irq 25 at device 1.1 on pci10
Oct  8 23:20:37 vpn1 kernel: em1: Ethernet address: 00:02:a5:4e:8d:b3
O

How-To-Repeat: Use both ports with 1000Mbit links.
Comment 1 Remko Lodder freebsd_committer freebsd_triage 2007-10-09 15:45:05 UTC 
State Changed
From-To: open->closed

duplicate of i386/117051
Comment 2 Pavel Stepanov 2007-10-09 16:56:05 UTC 
Hello bug-followup,

> From-To:        open->>closed
> By:   remko 
> When: Tue Oct 9 14:45:05 UTC 2007 
> Why:  duplicate of i386/117051

I cannot find PR such as i386/117051

This bug is still actual.

-- 
Best regards,
 Wishmaster                          mailto:wishmaster@velnet.ru
Comment 3 Mark Linimon freebsd_committer freebsd_triage 2007-10-10 00:52:51 UTC 
State Changed
From-To: closed->open

This PR was processed by both new freefall and old freefall (nosedive), 
and was thus falsely reported as a duplicate.  It should be re-opened here 
on new freefall.
Comment 4 Pavel Stepanov 2007-10-14 22:12:09 UTC 
Hello bug-followup,

  The problem is still here! Does anybody knows the solution?

-- 
Best regards,
 Wishmaster                          mailto:wishmaster@velnet.ru
Comment 5 Pavel Stepanov 2007-10-21 19:34:24 UTC 
Hello bug-followup,

 Checked with the latest Intel v. 6.6.6 .

 The same issue.

-- 
Best regards,
 Wishmaster                          mailto:wishmaster@velnet.ru
Comment 6 Mark Linimon freebsd_committer freebsd_triage 2008-01-07 06:18:40 UTC 
Responsible Changed
From-To: freebsd-i386->freebsd-bugs

This does not sound i386-specific.
Comment 7 Remko Lodder freebsd_committer freebsd_triage 2008-01-07 17:29:23 UTC 
Responsible Changed
From-To: freebsd-bugs->freebsd-net

Over to maintainer.
Comment 8 Vladimir Ivanov 2008-01-09 12:09:54 UTC 
Do you've IPMI board?

WBR,
Vladimir
Comment 9 Pavel Stepanov 2008-01-09 16:31:52 UTC 
Hello Vladimir,

Wednesday, January 9, 2008, 3:09:54 PM, you wrote:

VI> Do you've IPMI board?

VI> WBR,
VI> Vladimir


What do you mean under "IPMI board"?

I have server MB Asus  P5M2. (http://ru.asus.com/products.aspx?l1=9&l2=39&l3=352&l4=0&model=1396&modelmenu=1)

BTW, I've seen the same problem on another motherboards even for
socket 478 CPUs.

BTW1, There are no problems with the same network card but with the PCI-E 4x slot.
I think that the problems locates in PCI-X 64bit bus driver and not in
network card hardware or driver.

-- 
Best regards,
 Wishmaster                            mailto:wishmaster@velnet.ru
Comment 10 Vladimir Ivanov 2008-01-10 09:12:36 UTC 
Wishmaster wrote:
> What do you mean under "IPMI board"?

I've seen similar bugs if IPMI lan management has been installed.
Comment 11 Andre Oppermann freebsd_committer freebsd_triage 2010-08-23 19:13:11 UTC 
Responsible Changed
From-To: freebsd-net->jfv

Over to maintainer.
Comment 12 Sean Bruno freebsd_committer freebsd_triage 2015-06-29 18:31:14 UTC 
I do not see any changes to em(4) that would allow this to work now.  Can you retest this with current?
Comment 13 Sean Bruno freebsd_committer freebsd_triage 2015-08-03 16:52:13 UTC 
There are several handlers in the em(4) and lem(4) code for things like this.  I suspect they have been fixed at this time.